https://bugs.kde.org/show_bug.cgi?id=470874
--- Comment #3 from Vlad Zahorodnii <vlad.zahorod...@kde.org> --- (In reply to David Edmundson from comment #2) > Can reproduce. > > I believe the path is: > - client gets released (presumably unmapped?) > - we clear grouping > - they update an icon property > > A guard is trivial if that's the right direction. The assumption is that X11Window::windowEvent() won't be called if the window has been released or destroyed, so it lacks corresponding checks. -- You are receiving this mail because: You are watching all bug changes.
